International Business Machines Corporation

Harish Sukhwani

Hugh Edward Hockett

Giribabu V Paramkusham

Christopher M Laffoon

James C Riordan

Keith Rafferty

Nicholas Calvin Schambureck

An approach is provided in which a system displays a user interface that includes cloud environment configuration parameters on a display. Each of the cloud environment configuration parameters is selectable by a user and corresponds to a new cloud environment. The system receives a set of user selections from the user that selects at least one of the cloud environment configuration parameters. Then, the system generates a recommended cloud profile based, at least in part, on the set of user selections, and configures the new cloud environment based on the recommended cloud profile.

Configuration and usage pattern of a cloud environment based on iterative learning

An approach is provided in which a system receives a set of cloud requirements corresponding to a cloud environment and analyzes the set of cloud requirements using a set of machine logic based rules to generate a recommended cloud profile that is based, at least in part, on the set of cloud requirements and includes configuration setting values. In turn, the system configures configurable parameters of the cloud environment according to the configuration setting values of the recommended cloud profile.

Specific example debugging sessions are used to generate a generic decision tree (flow diagram) for use in future debugging sessions. Debug metadata is recorded for each debugging session. The debug metadata includes log file accesses and key traces. A generic decision tree is generated for use in future debugging sessions. Each file accessed by a developer during debug sessions creates a node in the generic decision tree. Branches of the decision tree are created between the nodes by aggregating the plurality of debug sessions together based on similarities and differences among the file accesses and results for the various debug sessions. The recorded metadata is genericized by identifying specific actions taken by developers and specific results thereof. The actions and results are abstracted to create more general types of actions and results such that the generic decision tree is applicable to a wider variety of problems in the software program.

Growing community of inventors

Durham, NC, United States of America

Harish Sukhwani